Selected Papers No. 24
l
The Meaning
of Business
Income: an
International
Comparison
By SIDNEY DAVIDSON
GRADUATE SCHOOL OF BUSINESS
UNIVERSITY OF CHICAGO
S IDNEY DAVIDSON is Arthur Young Professor of Accounting in The University of Chicago’s Graduate
School of Business, and Director of the School’s
Znstitute of Professional Accounting. The Institute
is in large part a product of Professor Davidson’s
work since he came to Chicago from Johns Hopkins
University in 1958. Davidson, a C.P.A., received
the A.B., M.B.A., and Ph.D. degrees from the University of Michigan. He has been a Visiting Professor at the Universities of Michigan, California, and
Hawaii, at Stanford, at the Hebrew University of
Jerusalem, and at the London School of Economics.
He is a member of the Accounting Principles Board
of the American Institute of CPA’s and was formerly
director of research of the American Accounting
Association. He also serves as a consultant to the
U.S. Treasury on taxation problems. His published
works include Studies in Accounting Theory (with
William T. Baxter) and An Income Approach to
Accounting Theory (with several colleagues on the
faculty of the Graduate School of Business). He is
the author of an earlier paper in this series-Selected
Papers No. 2, The Meaning of Depreciation. Basic
material for the research project which led to the
present Paper was gathered by Professor Davidson
during a tour of Western Europe in 1965. The
paper was initially presented as an Executive
Program Club Luncheon talk at the Pick-Congress
Hotel, Chicago, November 1, 1966.
The Meaning of Business Income:
an International Comparison
addressing this
group earlier in the year, said that Americans
have often given an extra measure of respect
to products of Western Europe-from English
shoes to Italian actresses. This respect may
arise from a variety of historical and cultural
factors. We are all humbled by the knowledge
that a portfolio of securities is no substitute
for a thousand years of culture when it comes
to distinguishing a Michelangelo from a
Titian, or arguing with a Parisian cab driver.
We may well ask whether the superior taste
and finesse necessary to deal with these situations will be applied in other spheres of activity as well-in, for example, the definition and
measurement of business income. English
shoes and Italian actresses may bulge in t h e
right places, but does European business income also bulge properly?
That this question is important-and here I
am referring to the question about business
income-is attested by many sources. A while
back we ran a survey of a group of businessmen and a similar one of graduate students in
business. We asked: “If you were a manager,
or an investor, or a potential investor, and
could have only one statistic about a given
firm, what statistic would you want?” Over 90
per cent of the respondents in both groups
said the statistic they would most want would
be the income of the business for the year. One
sage student said the statistic he would like
would be the income-for the next year; clearly, he’ll go far.
More seriously, I would like to ask how unambiguous, how universal is the meaning of
PROFESSOR ARNOLD WEBER,
2
this figure for business income? Does it mean
the same thing wherever we go? And if there
are differences, how do we explain them? How
have dissimilarities in economic, political, and
cultural factors in the several countries of
Western Europe produced different meanings
for business income? How do these differences
affect the financial statements, and, more importantly, the business actions that are based
upon these statements?
Two years ago John Kohlmeier and I
started on a research project to study these
questions; he deserves credit for most of what
follows except the faulty exposition.
Three Goals
Basically, we set out to do three things.
First, to discover the decision rules for determining business income in four Western
European countries-the United Kingdom,
France, The Netherlands, and Sweden. These
four were selected partly because they are
significant industrial and trading nations with
close ties to the United States, and partly because they typify major contrasting views of
the measurement of business income. The fact
that they are thoroughly delightful spots for
on-the-scene personal research was a relatively
trivial factor in their selection.
Second, we sought to discover how these
differences would affect t h e amounts reported
for income in two actual situations. To do this
we obtained the cooperation of two Chicagoarea manufacturing firms and their public
accountants, and decided to carry out t h e
analysis by a computer simulation. As Dean
Shultz has observed, for research to have
glamor nowadays it must use a computer.
What we did was subject the actual transactions for five years-1959 through 196%for
each of the two American companies to the
prevailing business income determination
rules of the European countries. What
3
emerged from the computer for each company
were five separate sets of five-year financial
statements. These statements were the actual
ones of the American firms, and the ones that
would have appeared if they had used the accounting practices-the rules for determination of business income-of each of the four
European countries. In other words, we had
five sets of statements showing how identical
transactions would have been recorded, and
how business income would have been reported, in each one of these five countries.
Finally, knowing the magnitude of these
differences we asked: What is likely to be the
effect of these differences on business decisions?
Let’s look at each of those points separately.
In determining the meaning of business income in each country, three major institutional sources were consulted: the tax laws,
the other relevant statutes, and the statements
of professional accounting organizations. In
the countries where the civil law concept is
well established-France and Sweden in our
example-tax laws play a dominant role in
determining business income procedure. In
these countries it is generally presumed that
items will be deductible for tax purposes only
if they are similarly reported in the financial
statements. In common law countries, tax laws
are less significant in determining accounting
principles, since tax and accounting reports
may, and frequently do, diverge.
A second source of business income determination rules are the statutes which specify
the requirements of annual reports for various
classes of companies-statutes similar to those
which apply to the Securities and Exchange
Commission in the United States. These
statutes are usually more concerned with the
form of financial reports than with the substance of valuation principles, but they do
4
have some effect on the reporting of business
income.
Finally, the recommendations by accounting
groups also help to determine accounting principles and the reporting of business income.
In none of the four Western European countries does an accounting organization have the
power of our Accounting Principles Board in
recommending principles and in requiring
disclosure of a failure to adhere to them, but
nevertheless the effect of ruling by these foreign accounting groups can be felt.
In addition to these institutional sources, individuals involved with the meaning of business income were interviewed in each country.
Included were business executives, officers of
accounting and business associations, representatives of international accounting firms,
and local accounting practitioners. These respondents were uniformly thoughtful and
courteous, and our indebtedness to them is in
no way diminished by the fact that the answers from different respondents within a
country were frequently contradictory.
It became clear that the definition of income within a country was rarely uniform and
so the procedures that follow are those which,
in our judgment, are the ones most commonly
used in each country-although others may
disagree.
English Procedures
Turning to individual countries: British
procedures for the determination of business
income have been, as you might imagine, evolutionary and permissive. As befits the mother
country of the common law, tax regulations
are not dominant in shaping business income
determination rules. As you may know, the
first United Kingdom income tax law was
enacted in 1842 as a temporary measure. The
British are never anxious to hurry things, and
April 1967 will see the 125th renewal of this
5
temporary legislation. Yet none of the 125 renewals have specified rules for the determination of business income; they have only set
out methods for determining taxable income,
and have implicitly contained a recognition
that business income and taxable income may
differ.
The Companies Acts in England, going back
to 1844, have always exerted some influence on
financial reporting. Even the most recent 1948
Act and the one that is proposed for action in
the current Parliament, however, have been
concerned primarily with the form and content of financial statements, and rather little
with the actual measurement of business income. The establishment of rules for the determination of business income, to the extent
that there is any persuasive force, has been left
largely to the Institute of Chartered Accountants in England; and although they have
tended to be permissive, their official Recommendations have been accorded substantial
weight.
Although practice is far from uniform, it
seemed to us that the traditional methods of
valuation of inventories and plant at acquisition cost, with a FIFO cost flow assumption
for inventories and a straight-line procedure
for depreciation, are most commonly used in
the United Kingdom.
French Methods
Turning to France: French financial reports
are largely determined by the provisions of
the “Plan Comptable Generale” a n d t h e
French tax law. The Plan was developed
by a government Committee for Accounting
Standardization in 1947 and has been revised
several times: the version now in use was
issued in 1962. It has been almost universally
adopted by publicly-held companies, which
explains the very substantial uniformity in
French financial reporting.
6
The valuation rules of the Plan generally
coincide with those found in the French income tax law; the tax regulations are virtually
controlling for financial statement valuation.
An unwritten rule in French accounting is
that you should never show more revenue or
lower expense in the financial statements than
in the tax returns. The force of this attitude is
indicated by the French reaction to the concept of deferred taxes-a peculiar American
innovation that says where a tax income and
business income differ you should recognize a
deferred tax liability. It is hard to explain and
justify in the United States and doesn’t export
easily. When I queried one preeminent
French practitioner about deferred taxes, he
replied that the concept was unheard of in
France. In fact, he went on to add that he
would not know how to express t h eidea in the
French language. I discussed this comment
with Professor William Vatter, the distinguished scholar who was on the accounting
faculty here for many years, and he observed
that it was one of the strongest arguments he
ever heard for making French the international language of accounting.
From 1945 to 1959 France, as you know, experienced a rapid inflation. During that time,
French tax laws permitted a variety of devices
that allowed firms to deduct the additional
cost of replacing inventory at inflated prices.
The last major revision of the t a x laws in
France occurred in 1959. This date coincides
with the devaluation of the franc, and prices
have been relatively stable since that time.
The 1959 revision of the tax law removed almost all of t h e provisions permitting firms to
adjust inventory expenses by the use of a general price level corrector.
The inflation had an even greater effect on
the accounting for plant assets than for inventory. Many successive revaluations had
been permitted for plant assets in France, and
7
all had been based on the application of a
general price level corrector to plant costs.
Even though the last revaluation was authorized by the 1959 tax law, it still affects pre-1960
property in use, and has had a substantial
effect upon the reporting of income in France.
So if we were to summarize French policy,
we would say that they adjust depreciation expenses for the changes in the general price
level-a procedure that also has been followed
to some degree in Belgium, Italy, and Japan.
The Netherlands
Turning to The Netherlands: there has
probably been greater freedom in development of concepts of business income there
than in any other country that we studied;
neither the tax laws nor the Dutch Code of
Commerce imposes limitations on what may be
done in financial reporting. Thus the accounting profession has been free to make whatever
adjustments in the reporting of income its
members thought appropriate; and this freedom to develop principles has been exercised
by a professional accounting community nurtured by extremely close connections with the
Dutch academic institutions. The ties have
been especially close with the university group
of business economists, and-as we might anticipate-there has been a consequent focusing
on replacement values in the determination of
expenses and income. This has introduced a
substantial judgmental element into the determination of income, but Dutch financial reports seem to be highly regarded by all groups
-domestic and foreign.
In Sweden, income determination and financial reporting are probably more affected
by national economic goals than in any other
country. Efforts to smooth the swings of the
business cycle dominate Swedish national economic policy, and these efforts apply with special force to tax policy. The linkage with in-
come determination comes from the civil law
rule that taxable income will be determined
from the taxpayers’ financial records. In general, deductions can be claimed on the tax return only if they also appear in the financial
reports.
How does this affect our major areas? Let
us consider two-inventories and depreciation.
Inventory accounting is completely dominated
by the tax law; its basic provision permits a
taxpayer to write down inventory to 40 per
cent of cost or market, whichever is lower.
Thus he can write inventory down by 60 per
cent of cost, but the deduction from taxable
income is permitted only if it also appears as
an expense in the financial reports.
Accounting for depreciation compares in
liberality and flexibility with inventory accounting. The firm has its choice of two
methods: The first provides a 30 per cent declining balance rule-that is, you can write off
30 per cent of the undepreciated balance. The
second alternative is a 20 per cent straight-line
provision. Both alternatives have the effect of
allowing the firm to write off all of its plant
over a five-year period on its tax return if it
also does in its financial records. In fact, if it
doesn’t take the full 20 per cent one year, it
can take more than 20 per cent the next.
In addition to these, and as a special feature
of the efforts to smooth the swings of the business cycle, the Swedish tax law provides for a
system of investment reserves; corporations
may deduct an amount equal to 40 per cent of
their pre-tax income in the computation of
taxable income as an investment reserve. That
is, the firm figures its pre-tax income, then can
deduct up to 40 per cent of that amount as an
investment reserve; the remainder is subject to
the income tax. But of course this can be done
only if the investment reserve expense is
shown as a deduction in the calculation of
business income. If the investment reserve ex-
9
pense is claimed, an amount equal to 46 per
cent (approximately the effective corporate
tax rate) of the investment reserve must be
deposited in a noninterest-bearing bank account in the state bank. Later, with the approval of the State Labor Board, the firm can
withdraw the funds from this account to pay
for new plant and equipment. Approval by
the Labor Board of expenditure of investment
reserve funds is likely to come only in periods
of depressed business conditions.
It is apparent that the determination of
business income in Sweden is a very flexible
sort of thing. In fact, one Swedish businessman
told me that at the end of the fiscal period the
directors decide how much profit to report. He
said they consider such things as the effect the
reported profit may have on forthcoming wage
negotiations, t h e amount of dividends they
want to pay, and a proper relation between
dividends and income; and when the directors
have determined approximately what profit
they want to report, they tell the accountant
to find the easiest way to adjust the books to
achieve the desired figure.
That was the businessman’s way of explaining income determination in Sweden; a Swedish accountant explained it somewhat differently. He said, “In Sweden we start at the top
and bottom of the income statement and work
toward the middle, minimizing taxes along the
way.” Sales are determined by outside forces,
and the directors decide on the size of the dividend they want to pay, so you get the top and
the bottom. “Then,” the accountant went on to
say, “we determine the expenses to fit. The
larger the reported income, the more income
tax, so we maximize reported expenses, within
t h e provisions of the law and recognizing the
amount of dividends we want to pay.”
How, then, shall we characterize the mean-
10
ing of business income in these four countries?
The British approach to reporting business income is staid, traditional, but with results that
are not conservative; the French approach is
uniform and relies on general price level adjustments; the Dutch introduce replacement
cost concepts and get respected judgmental reports of earnings; the Swedes make the reporting of business income serve the national economic goals of stability and growth, so that
the concept of business income becomes almost
unrecognizable.
Before turning to the measurement of the
effect of these differences, it probably is necessary to note that there are also variations in
the American definition of business income. In
the United States, although we stick with the
acquisition cost concept in inventory accounting, we do have a choice between LIFO (Last
In, First Out) and FIFO (First In, First Out)
flows; we also have a choice between straightline and accelerated depreciation.
Measuring the Effect
To return to the question: What effect do
the differences among the accounting conventions of the countries involved have on reported business income? To answer this question, as I said earlier, we selected two Chicago
firms as case studies. One of the two, a steel
products producer, was very stable with little
movement in sales or income during the five
years studied-1959-63. Sales for the period
varied between $4.2 and $4.7 million per year;
net income as reported varied from $190,000
to $240,000 over the period, with no special
growth factors.
We wanted to see whether the effect of differences in the definition of business income
was influenced by growth, so our second company-an electric products manufacturer-was
selected because of its rapid growth in the five
years preceding 1963. In that period sales ad-
11
vanced from slightly over $10 million to well
over $15 million, and reported income more
than doubled, from a little over $250,000 to
well over $600,000.
For each company, we summarized external
transactions during the five-year period and
converted these transactions into computerusable form. The procedures actually used by
each company were also translated into t h e
computer program, and then permitted to
operate on the transaction data.
After more frustrations than you have patience to listen to, we finally obtained printouts of the financial statements of each firm
that were within 2 per cent on all items of
their actual financial statements, with most
items being exactly reproduced.
Having established the validity of our
model, we then put into the computer the set
of decision rules for determining business income in each of the four foreign countries,
and allowed these rules to operate on actual
transactions of the U.S. firms. Twenty seconds
later we had a complete set of financial statements for each U.S. company-that is, a balance sheet, an income statement, a cost-ofgoods-sold schedule, and a statement of fund
flows, both in dollars and as percentages of
U.S. numbers. We had a series of financial
statements-five different ones for each company, each series covering five years. We had,
as you might infer, a mountain of data about
each of these two companies. I will sketch here
only in the very briefest terms what the data
show; a fuller summary appears in the Autumn, 1966, issue of The Journal of Accounting Research.*
Although our data contain detailed analyses
of cost of goods sold and depreciation expense,
let me focus on net income, and because of
*“A Measurement of the Impact of Some Foreign
Accounting Principles,” by Sidney Davidson and John
M. Kohlmeier.
12
difficulties of approximating taxes in the foreign countries, I’ll talk almost exclusively
about net income before taxes.
The U.S. companies both used straight-line
depreciation for financial reporting, although
they did use accelerated depreciation for tax
returns. The electric company used FIFO inventories, the steel company used LIFO, but
to get comparability between them we converted the steel company data to a FIFO system. Hence we are talking about the U.S. companies using straight-line depreciation and
FIFO inventory determination.
When we turn to the numbers, one fact
stands out most clearly: The combination of
FIFO and straight-line depreciation in the
U.S. produces pre-tax income figures as large
or larger than any other set. There were 40
figures for net income before taxes for the
other countries and 38 of those were less than
the comparable figure for the U.S. That is, we
had four different countries, two companies,
and five years, so we had 40 net income figures,
other than the U.S. figures, and 38 figures of
those 40 were incomes lower than were reported in the U.S.-and the two that weren’t
lower were exactly equal to them.
Swedish income figures were always the
lowest of the group: even excluding Sweden,
the pre-tax incomes in the other countries
were between 81 per cent and 99 per cent of
the U.S. figures for the electric company and
between 87 per cent and 100 per cent of the
U.S. figures for the steel products company. As
we surmised, the growth of the electric products company meant that there was a greater
deviation in its reported income as determined
by foreign standards and U.S. standards than
was true for the more stable steel products
company. Swedish figures were far below all
the others, varying for the electric products
company from 19 per cent of U.S. income in
one year to 59 per cent in another; for the
steel company the Swedish results vary between 50 per cent and 59 per cent.
One other point that becomes clear from
this analysis is that the substitution of LIFO
or accelerated depreciation in the U.S. tends
to produce variations from FIFO and straightline that are approximately as great as any
shown by any foreign country’s accounting
principles, other than the Swedish ones. That
is, if we take permissible alternatives within
the United States they’ll give us as great a difference in reported income as we get by substituting the financial principles of any country other than Sweden.
Summary of Findings
What can we say to summarize the findings?
Let me point out first that any conclusions to
be drawn start with the warning that the data
relate only to the reports of two firms for five
years; and while we noted no unique features
about these firms, neither would we claim that
they are typical-if there is such a thing as a
typical firm. Also, the five-year period studied
was one of price stability; it would be hazardous to predict t h e results of other companies
for the same years, or these companies for
other years, and it would be especially hazardous to predict the results for other companies
for other years. Nevertheless, once this model
is developed and the computer program is prepared, t h e inputs of data for other companies
and other years can be handled with little difficulty. We are now moving ahead on this
project, in the hope that we can extend the
analysis and obtain additional generalizations.
But t h e one primary generalization that these
data do provide at the moment is that U.S.
straight-line and FIFO income figures are
higher than those reported anywhere else.
The data also suggest that stability of operations and prices tends to narrow the effects of
differences in accounting procedures. In a per-
14
fectly stable world it doesn’t matter too much
what business income rules you use, and this
we validated by putting in some estimated
price level changes. That is, we sought to find
what would happen if prices had been rising
during this period instead of being stable, and
we got some fairly substantial differences.
A third generalization that I think is warranted is that the wide departure of t h eSwedish figures from those of any other country,
and their erratic movements, emphasize the
danger of linking financial statements to t a x
reports, especially when tax provisions are an
important element in the national economic
plan. We have avoided that danger in the
United States, except with regard to LIFO. I
think the lesson to be learned is that when
financial statements and tax returns must coincide, reported business income becomes the
prisoner of the tax law and meaningful statements of business operations are destroyed.
Finally, there remains the question of how
these differences in financial reporting affect
individual investment decisions and the aggregate flow of capital. The unsatisfactory state
of our present knowledge about the basis of
investment decisions and capital movements
prevents us from drawing any general conclusion. It is likely, though, that assessments of
risk and income as evidenced by financial
statements play some role in these decisions,
perhaps a relatively important one. And if
this is so, some quantitative measure of the
effects of international differences in accounting principles may be helpful in dealing adequately with this most important problem.